Genital Chlamydia trachomatis Infections Clear More Slowly in Men Than Women, but Are Less Likely to Become Established

Abstract

Background: Rigorous estimates for clearance rates of untreated chlamydia infections are important for understanding chlamydia epidemiology and designing control interventions, but were previously only available for women.

Methods: We used data from published studies of chlamydia-infected men who were retested at a later date without having received treatment. Our analysis allowed new infections to take one of 1, 2, or 3 courses, each clearing at a different rate. We determined which of these 3 models had the most empirical support.

Results: The best-fitting model had 2 courses of infection in men, as was previously found for women: "slow-clearing" and "fast-clearing." Only 68% (57%-78%) (posterior median and 95% credible interval [CrI]) of incident infections in men were slow-clearing, vs 77% (69%-84%) in women. The slow clearance rate in men (based on 6 months' follow-up) was 0.35 (.05-1.15) year-1 (posterior median and 95% CrI), corresponding to mean infection duration 2.84 (.87-18.79) years. This compares to 1.35 (1.13-1.63) years in women.

Conclusions: Our estimated clearance rate is slower than previously assumed. Fewer infections become established in men than women but once established, they clear more slowly. This study provides an improved description of chlamydia's natural history to inform public health decision making. We describe how further data collection could reduce uncertainty in estimates.

Keywords: Bayesian inference; chlamydia; evidence synthesis; natural history; sexually transmitted diseases.

Figure 1.

Validation of the model, showing crude (data; open circles) and simulated (model; filled circles) clearance rate for each data point in Table 1. Error bars give the 95% confidence interval for crude estimates, and the central 95% of simulations. For crude estimates, an arrow pointing right indicates that all men cleared the infection, so there is no upper bound or estimate for the crude clearance rate (see also the legend to Table 1). The first 6 studies (above the dashed line) were clinic-based, and the last 2 (below the line) were screening studies. Within this grouping, studies are ordered by maximum duration of follow-up. Follow-up in years at each observation is indicated on the left-hand side.

Figure 2.

A, Simulated survival curves for incident and prevalent infections in men. Solid lines indicate the median of the simulated proportion persisting, shaded areas the central 50% (the interquartile range), and dashed lines the central 95%. Note that the time axis is time since infection for incident infections, and time since detection of infection for prevalent infections. B, Posterior distributions for the clearance rate of slow-clearing chlamydia infections in men (blue) and women (pink), based on follow-up of up to 6 months in men and 4 years in women.

Figure 3.

Posterior distribution for the difference between parameters characterizing infections in men and women. The x-axis represents differences in clearance rate and the y-axis represents differences in the proportion of infections that are slow-clearing. Solid lines at x = 0 and y = 0 divide samples in which parameters are higher/lower in men vs women.